Iron-Mediated Lysosomal Membrane Permeabilization in Ethanol-Induced Hepatic Oxidative Damage and Apoptosis: Protective Effects of Quercetin

Iron, in its free ferrous states, can catalyze Fenton reaction to produce OH∙, which is recognized as a crucial role in the pathogenesis of alcoholic liver diseases (ALD). As a result of continuous decomposition of iron-containing compounds, lysosomes contain a pool of redox-active iron. To investigate the important role of intralysosomal iron in alcoholic liver injury and the potential protection of quercetin, male C57BL/6J mice fed by Lieber De Carli diets containing ethanol (30% of total calories) were cotreated by quercetin or deferoxamine (DFO) for 15 weeks and ethanol-incubated mice primary hepatocytes were pretreated with FeCl3, DFO, and bafilomycin A1 at their optimal concentrations and exposure times. Chronic ethanol consumption caused an evident increase in lysosomal redox-active iron accompanying sustained oxidative damage. Iron-mediated ROS could trigger lysosomal membrane permeabilization (LMP) and subsequent mitochondria apoptosis. The hepatotoxicity was attenuated by reducing lysosomal iron while being exacerbated by escalating lysosomal iron. Quercetin substantially alleviated the alcoholic liver oxidative damage and apoptosis by decreasing lysosome iron and ameliorating iron-mediated LMP, which provided a new prospective of the use of quercetin against ALD.


Introduction
Alcoholic liver disease (ALD) is a primary consequence of heavy and prolonged drinking and represents a major cause of morbidity and mortality worldwide [1]. It is now well accepted that the pathogenesis of ALD is multifactorial, and iron-mediated oxidative stress has attracted emerging attention [2]. Alcohol stimulates iron absorption and hepatic storage by inhibiting hepcidin secretion, which initiates "free" iron release from various iron-containing proteins [3]. Cellular "free" iron, present as labile iron pool (LIP), is low-molecular weight, redox-active, and weakly chelated iron [4]. In case of excess, such redox-active isoform of iron may synergistically and dramatically amplify ethanolinduced oxidative stress and then exacerbate alcoholic liver injury by serving as a catalyst for hydroxyl radical ( • OH, the most toxic fraction among various reactive oxygen species (ROS)) formation via Fenton/Haber-Weiss reaction [5].
Lysosomes are responsible for autophagic turnover of all organelles and most long-lived proteins. As a consequence of degradation of many iron-containing macromolecules (e.g., ferritin and mitochondrial components), lysosomal 2 Oxidative Medicine and Cellular Longevity compartments can accumulate large amounts of low mass redox-active iron and, therefore, enhance susceptibility to oxidative stress [6,7]. Moreover, the ongoing digestion of such metalloproteins is accompanied by the release of redoxactive iron which, upon export from the lysosome, may be a major intracellular source of LIP. In our previous work, we considered that mucolipin 1 mediated the release of endosomes and lysosome iron contributed to the increased LIP induced by ethanol [8]. Additionally, lysosomes are intimately associated with apoptotic cell death. Iron-mediated lysosomal ROS generation causes lysosomal membrane permeabilization (LMP) [9] and the ensuing release of cathepsins into the cytosol and thereby induces apoptosis through activation of mitochondrial membrane permeabilization (MMP) [10]. Growing studies have demonstrated that intralysosomal iron is a key determinant of ROS-induced injury [11,12], and LMP has been presumed as an early event of lysosomemitochondrial apoptosis pathway based on the blocking effect of deferoxamine (DFO) which is a typical lysosomal targeted iron chelator [13]. However, it still remained unclear, to our knowledge, what the role of intralysosomal redoxactive iron is upon ethanol-induced hepatic oxidative damage and apoptosis.
In spite of common use for iron chelation therapy, DFO exerts the main limitation because of its potential toxicity and parenteral administration by slow subcutaneous injection with poor patient compliance [14]. Thus, there is an accumulating interest in the natural iron chelators originated from dietary phytochemicals. Quercetin, a naturally occurring bioflavonoid widely distributed in vegetables, fruits, and herb medicine, scavenges free radicals and exhibits ironbinding activity dependent on its structure characterized as phenolic hydroxyl groups [15]. We found that quercetin evidently attenuates ethanol-derived damage as a result of its prominent antioxidative properties [16]. Quercetin has been also demonstrated to significantly decrease hepatic iron and effectively quench intracellular free iron-induced • OH production [17]. Furthermore, quercetin supplementation modulated lysosomal enzyme activities in various diseases [18,19] suggesting that the hepatoprotective effect of quercetin against alcohol-derived injury might be involved in the regulation and chelation of quercetin on lysosomal labile iron. Therefore, we hypothesized that lysosome redoxactive iron may play an important role in ethanol-induced liver damage and quercetin potentially relieves ethanolelicited hepatic oxidative damage and apoptosis by decreasing intralysosomal redox-active iron-mediated ROS overproduction and the subsequent LMP. Hence, we designed this study to probe the role of intralysosomal low mass iron toxicity in the process of ALD and the protection of quercetin.

Animal Treatment.
Animals were cared for according to the Guiding Principles in the Care and Use of Laboratory Animals published by the US National Institutes of Health. The male C57BL/6J mice (18∼20 g), obtained from Sino-British Sippr/BK (Shanghai, China), were pair-fed with either regular or ethanol-containing Lieber De Carli liquids diets (Beijing HFK Bioscience Co., Ltd., Beijing, China) for 15 weeks. Following 1 week of acclimation, the mice were divided randomly into six groups of fifteen each: (1) normal control group (C), (2) ethanol group (E, 30% of total calories as ethanol), (3) ethanol plus quercetin group (EQ, quercetin: 100 mg/kg.bw), (4) quercetin control group (Q), (5) ethanol plus DFO group (ED, DFO: 100 mg/kg.bw), (6) DFO control group (D). Quercetin and DFO were received by gavage and intraperitoneal injections, respectively, according to the body weight. After 15 weeks, hepatocytes were isolated from three mice of each group to assay lysosomal status and intralysosomal iron content. Other mice were sacrificed after an overnight fasting and hepatic samples were collected rapidly for further biochemical assays.

Primary Hepatocytes Isolation and Treatment.
According to a two-step collagenase described by Figliomeni and Abdel-Rahman [20], primary hepatocytes were isolated from adult male C57BL/6J mice (22∼25 g). The freshly harvested hepatocytes (≥90% viability) were suspended in DMEM medium with 10% fetal bovine serum, 100 U/mL penicillin, 100 U/mL streptomycin, 32 IE/L insulin, 0.1 mol/L hydrocortisone, and 15 mol/L HEPES and then inoculated in rat tail collagen-coated 6-well plates for culture in a humidified incubator at 37 ∘ C, 5% CO 2 . On the following day, the attached hepatocytes were incubated with medium containing various pharmacological reagents. Before subjecting to ethanol (100 mmol/L), some cultures were exposed to bafilomycin A1 (100 nmol/L), FeCl 3 (60 mol/L), DFO (1 mmol/L), and quercetin (100 mol/L). After 24 h, the supernatant and cells were collected for various bioassays according to the corresponding experimental protocol. The optimal concentration and exposure times of the all regents were established in preliminary experiments.

Measurement of Hepatic Redox Status.
Malondialdehyde (MDA) was measured by using thiobarbituric acid colorimetry slightly modified by Ohkawa et al. [21]. Reduced glutathione (GSH) and oxidized glutathione (GSSH) were measured with the assay kit. The MDA level was standardized by protein concentration evaluated with the Bio-Rad protein assay kit (USA).
DHE was used for in situ determination of ROS in mice liver tissues. The frozen hepatic sections (5 m) were immediately incubated with DHE (5 M) at 37 ∘ C for 15 min. Then the red fluorescence intensity excited by green laser was visualized with a fluorescence microscope (Melville, NY). Fluorescence intensities in randomly selected areas of the images were quantified using IPP 6.0 image analysis software. ROS production in mice primary hepatocytes was assayed by adding H 2 DCFDA (10 M) to the culture medium for 30 min, followed by the quantifications with spectrometry (Ex: 485 nm, Em: 525 nm).

Apoptosis Assay.
Apoptotic cells were identified in formalin-fixed, paraffin-embedded liver tissue sections, with the classic in situ Dead End TM fluorometric TdT-mediated dUTP nick end labeling (TUNEL) assay according to the manufacturer's protocol. The expression of cleaved caspase-3 was measured by Western blot and caspase-3 activity was detected with the assay kit (Beyotime, China) according to manufacturer's instructions. In cell experiment, the apoptosis cells were measured by flow cytometryusing annexin V/PI assay kit.

Assay of Hepatic Total Iron, LIP, and Lysosomal Low Mass
Reactive Iron. Total amount of hepatic iron was determined by flame atomic absorption spectrometry (spectrAA-240FS, USA) using an iron hollow cathode lamp. The analysis of LIP was performed according to Khan et al. [22]. In brief, liver homogenates (2.5%) were prepared using 1 mM ethylenediaminetetraacetic acid (EDTA) to dissociate LIP and centrifuged at 20,000 ×g for 15 min at 4 ∘ C. The resultant supernatants were ultra-filtrated on Micron-30 at 14,000 ×g for 20 min at 4 ∘ C for atomic absorption spectrophotometric assay as LIP.
For detection of lysosomal low mass reactive iron content, we used the modified sulfide-silver method (SSM) [23,24]. In brief, hepatocytes were grown on coverslips and were then washed with PBS and fixed with 2% glutaraldehyde in 0.1 M Na-cacodylate (Amresco, USA) buffer with 0.1 M sucrose (pH = 7.2) for 2 h at 22 ∘ C. Next cells were sulfidated at pH ≈ 9 with 1% (w/v) ammonium sulphide (Zhenxing, China) in 70% (v/v) ethanol for 15 min. The development was then performed in the dark at 26 ∘ C for 20∼50 min using a physical, colloid-protected developer containing silverlactate (Sigma, USA). Following hydroquinone in a graded series of ethanol solutions and mounting in Canada balsam (Ourchem, China), the cells were examined with microscope.

Vital Imaging of Lysosomal ROS.
Hepatocytes were coincubated with H 2 DCFDA (10 M) and LysoTracker Red (100 nM) for 30 min, and images were sequentially collected in the green and red channels with the fluorescence microscope and merged with IPP 6.0 software.

Analysis of Lysosomal Membrane Permeabilization (LMP).
LMP was evaluated by acridine orange (AO), a lysosomotropic base containing metachromatic fluorophore, which becomes charged and retained within acidic compartments [25]. Hepatocytes were stained with AO (5 g/mL) in complete medium at 37 ∘ C for 15 min. AO-induced red (lysosomal) and green (nuclear and cytosolic) fluorescence were observed using fluorescence microscope (Melville, NY). LMP was also measured by flow cytometry. The green (FL1) and red (FL3) fluorescence of 10,000 cells was recorded on a logarithmic scale using a Becton Dickinson FACScan instrument while excited at 488 nm (argon laser). Using this technique, the alterations of lysosomal stability were assayed, as evident by decrease in FL3 red fluorescence. With the alteration of LMP, we measured lysosomal cathepsin B relocation and activity by Western blot and assay kit, respectively.

Analysis of Mitochondrion Membrane Permeabilization (MMP).
Cells were incubated with 100 nM TMRM for 30 min at room temperature and washed with PBS, and red mitochondrial fluorescence of 10,000 cells per sample was determined by flow cytometry by using the FL3 channel. Cells with increased MMP showed decreased TMRM red fluorescence.

Western Blot Analysis.
Mice liver tissues were homogenized and lysed at 4 ∘ C in RIPA buffer and the lysates with equal amount of proteins were subjected to Western blot. The protein was separated by 10% SDS-polyacrylamide gel and then transferred onto a PVDF membrane. After incubation with the specific primary antibodies against the target protein and the species-specific second antibodies conjugated to horseradish peroxidase, immunoreactive bands were detected by means of an ECL plus Western Blotting Detection System (Amersham Biosciences, Little Chalford, UK). The band densities were measured by Image J software.

Statistical Analysis.
Results are given as the mean ± standard deviation (SD) from replicate experiments and subjected to one-way ANOVA followed by Student-Newman-Keuls multiple range test (SPSS 16.0 software package). The statistical significance was defined as a difference between groups of < 0.05. Table 1, neither quercetin nor DFO had influence on final body weight and liver ratio against body weight (liver index) in comparison with normal control at the end of experimental periods. However, the weight gain of all ethanol-fed mice was lower than those of non-ethanol-challenged groups, and ethanol intake led to the marked increase in the liver index, whereas quercetin suppressed these adverse effects.

The Effect of Quercetin Supplementation on Body Weight and Liver Index of Mice Fed with Ethanol. As shown in
Interestingly, DFO had no influence on mice body weight gain compared with ethanol group, while DFO normalized the increased liver index delivered by ethanol.

Quercetin and DFO Attenuated the Mice Liver Oxidative Damage Induced by Chronic Ethanol Consumption.
As depicted in Figure 1, long-time ethanol treatment substantially provoked mice liver oxidative damage manifested by a remarkable MDA elevation, GSH depletion, and overproduction of ROS. As expected, both quercetin and DFO evidently normalized GSH level, decreased MDA content, and suppressed ROS generation derived from ethanol. However, quercetin and DFO per se had no influence on hepatic redox status compared with normal mice.

The Protective Effect of Quercetin and DFO against
Ethanol-Induced Hepatic Apoptosis. In situ apoptosis in mice liver section was evaluated by TUNEL staining. In the control group, quercetin group, and DFO group, TUNELpositive cells were not observed. However, the number and signal density of TUNEL-positive hepatocytes significantly increased in the ethanol-exposed mice (Figures 2(a) and  2(b)). Enhanced protein of the apoptotic marker cleaved caspase-3 confirmed this finding, with a 4.8-fold increase in ethanol group compared to normal control (Figure 2(c)). In addition, long-time ethanol consumption increased caspase-3 activity (Figure 2(d)). Quercetin and DFO treatment apparently reduced the reactivity and the number of apoptotic hepatocytes compared with the ethanol group. And the expression of cleaved caspase-3 and caspase-3 activity were significantly decreased in quercetin and DFO treatment group.

Quercetin, Similar to DFO, Suppressed the Ethanol-Induced Iron Abnormity of Hepatic Tissue and Especially
Lysosomal Compartment. The hepatic contents of total iron and LIP were measured by flame atomic absorption spectrophotometry. As illustrated in Figures 3(a) and 3(b), chronic ethanol feeding gave rise to intracellular accumulation of iron in liver. 1.4-fold and 1.5-fold increase of liver total iron content and LIP compared to normal control were observed. And the expression of ferritin light chain (Ft-L) was upregulated (Figure 3(c)). Further, the intralysosomal redox-active iron was determined using SSM. Compared with normal control, the intralysosomal iron (Figures 4(a) and 4(b)) was significantly increased in ethanol-exposed mice with a distinct lysosomal silver precipitate accompanying the high iron content and hepatic LIP. The abnormal increases of both hepatic and lysosomal iron were partially reversed by quercetin or DFO supplementation. Interestingly, quercetin per se had no influence on hepatic total iron content, LIP, and lysosome iron level, while DFO treatment resulted in manifest decrease of liver iron comparing to normal control. Concomitant with the increase of intralysosomal redox-active iron, ROS largely colocalized with lysosomes. Quercetin decreased the lysosomal ROS generation as well as DFO (Figure 4(c)). were incubated with bafilomycin A1, FeCl 3 , and DFO before exposing to ethanol. The intralysosomal reactive iron content and lysosome status were measured by SSM and AO-uptake technique, together with cellular injury assay. As shown in Figure 6, compared with normal intact hepatocytes, ethanol markedly increased lysosomal silver precipitation, indicting a substantial accumulation of lysosomal redox-active iron. Moreover, ethanol incubation decreased the lysosome stability evidenced by lowered red fluorescence of AO, caused ROS overproduction, increased MMP, caused apoptosis, and elevated the release of cellular AST and LDH. What is worse, the intralysosomal iron level of the hepatocytes treated with FeCl 3 was higher than the ethanol group, and the excess iron increased further lysosomal destabilization and hepatic damage. This damage was blocked by bafilomycin A1 (bafilomycin A1 is a highly potent and selective vacuolar type H + -ATPase (V-ATPase) inhibitor that inhibits the acidification of lysosomes, thus blocking lysosome activity and degradation) or DFO (chelation of redox-active iron) by reducing the lysosome low mass reactive iron.

Discussion
Fenton reaction, a homolytic splitting of H 2 O 2 yielding the extremely reactive hydroxyl radical ( • OH), may mainly take place inside the lysosomal compartment where iron occurs in its low mass redox-active form (be regarded as lysosome LIP) [27]. • OH mediated cell damage may be initiated by peroxidative injury to lysosomal membranes. Resulting relocation of low mass iron and cathepsins would lead to secondary harm to various cellular constituents. There were substantial evidences for the major determinant of lysosomal iron induced oxidative damage and apoptosis in various diseases including neurodegenerative diseases, atherosclerosis, and diabetes [28][29][30]. In our current study, chronic ethanol abuse caused intralysosomal iron accumulation concomitant with the increase of hepatic total iron and ferritin light chain. Consequently, liver oxidative damage and lysosomemitochondrial apoptosis were triggered. Such detrimental outcomes were notably reversed by quercetin and DFO. Thus, the role of aberrant lysosome LIP level and lysosome damage in alcoholic liver injury was a focus of our investigations since iron chelation especially the intralysosomal redoxactive iron chelation was assumed as a new plausible strategy to ameliorate ALD. DFO, the common lysosomal target iron chelation, may accumulate within lysosomes and cause iron starvation with further negative consequences for iron metabolism [31,32]. In this work, as a result of long-time DFO intraperitoneal injection, the mice liver iron was significantly lower than normal control, which may disturb liver iron metabolism. Excitingly, plant-derived quercetin was shown to be a potent iron chelator similarly active to DFO, which, at least partly, contributes to its antioxidative properties [15,33]. Moreover, quercetin could readily permeate cell membranes via glucose transport proteins (GLUTs) [34] to quench intracellular • OH as a shuttle for labile iron [35,36]. As we expected, quercetin manifested an extensive protection against ethanol stimulated abnormal intralysosomal LIP level and the ensuing liver oxidative damage and apoptosis. It has been proposed that ferritin-iron release requires lysosomal activity, which is a major source of lysosome redox-active iron [37]. Ferritin is a main cellular iron-binding protein and its light chains (Ft-L) were mainly expressed in liver which may contain large amounts of iron [38]. Our data showed that Ft-L expression was obviously upregulated by ethanol accompanying the high hepatic iron content, to some extent, by a self-protection mechanism. Moreover, some studies indicated that ferritin seems to be capable of binding intralysosomal iron temporarily to protect against iron-mediated oxidation when it is autophagocytosed [11]. In spite of this, we still believed that the recycling of ferritin-iron in lysosomal compartment was the major source of intralysosomal redox-active iron since the iron-binding capacity of ferritin was weak in acidic lysosomal lumens [39]. Quercetin partially normalized ferritin expression, which may contribute to its hepatoprotective effect by reducing intralysosomal redox-active iron. LMP is considered to be a main mechanism of lysosome iron-induced damage [40]. Induction of LMP by ROS production has been extensively studied in model systems in which the intralysosomal iron content has been manipulated [41]. In our study, with the increasing lysosome iron, longtime ethanol consumption initiated ROS overproduction especially in lysosomes (colocalization with LysoTracker) and trigged serious LMP. It is admitted that LMP can cause iron release to cytoplasm directly and consequent prooxidant cell damage [42,43]. What is worse, other lysosomal contents (e.g., H + , proteases) also leak which in turn accelerates the labile iron release from cytoplasmic ferritin to LIP [7,39]. Here, our data showed that ethanol-elicited hepatic abnormal LIP was blocked by quercetin and we attributed repairing LMP to the positive effect. Evidence showed that the efflux of iron from the UVA-damaged lysosome was a major source of LIP in cells and epicatechin (one of the flavonoids) suppressed UVA-mediated release of labile iron by alleviating lysosomal rupture and ferritin degradation [44,45]. Besides, LMP is emerging as an important regulator of cell apoptosis. LMP allows the release of cathepsin B, an apoptotic mediator, into the cytoplasm, where it can initiate the intrinsic apoptotic pathway [46]. Under normal physiological condition, cathepsin B is sequestered into the lysosomes of intact cells to participate in normal turnover of proteins [47]. Once into the cytoplasm, cathepsin B can cleave to activate proapoptotic proteins, including Bid [48], followed by engaging MMP and release of cytochrome C [49]. Thereafter, the release of cytochrome C causes the activation of effector caspases and triggers a caspase-dependent apoptotic pathway [50]. Our results suggested that chronic ethanol consumption triggered lysosome-mitochondrial apoptosis pathway. Quercetin quenched iron incited lysosomal ROS and blocked LMP. Subsequently, quercetin not only inhibited the leakage of cathepsin B into cytoplasm but arrested all measured parameters of mitochondrial apoptosis pathway including MMP, the escape of cytochrome C from mitochondria, and caspase-3 activity.
To further probe the pathophysiological role of lysosomal iron in ethanol triggered hepatic injury and the protection of quercetin, we utilized some strategies to regulate intralysosomal redox-active iron pool. In neutral medium, FeCl 3 could form insoluble iron complexes and enter the lysosomal compartment by endocytosis [12]. We identified that enhancing free lysosome iron by adding FeCl 3 aggravated ethanol stimulated LMP and subsequently sharpened ROSinduced damage and MMP-mediated apoptosis. Such harmful results, conversely, were counteracted by lysosomal iron chelator DFO. Furthermore, preincubation with bafilomycin A1 to block lysosome function [10] ameliorated primary hepatocytes damage, suggesting that ethanol-induced hepatic oxidative injury and mitochondria apoptosis involved low mass iron-delivered lysosome rupture. Most importantly, the detrimental effects on hepatocytes triggered by ethanol were also hampered by quercetin supplementation. Renovating LMP accompanied with substantially decreasing hepatic intralysosomal redox-active iron is more likely explanation of the powerfully protective activity of quercetin. To our knowledge, it is the first time to report that redox-active lysosomal iron seemed to be chief reason for ALD development. Overall, it may be a new potential strategy to attenuate alcoholic liver injury by reducing lysosomal redox-active iron and the ensuing LMP.

Conclusion
All these findings strongly support the central role of a high amount of redox-active lysosomal iron in ethanol-elicited liver oxidative damage and apoptosis. Lowering intralysosomal redox-active iron pool and consequently the mitigation of LMP may be the potential mechanism of the powerful protection against alcoholic liver damage of quercetin. Our overall results highlighted a novel potential prospective for the prevention of ALD by naturally occurring iron-chelating quercetin.

Conflict of Interests
The authors declare no conflict of interests.